Dominance interactions, spatial dynamics and emergent reciprocity in a virtual world

In many studies on social demeanour it is overlooked how complex social behaviour may result from simple self-reinforcing interactions between entities. In this paper an individualoriented computer model is used to study whether the spatial dynamics between artificial agents and their autocatalytic interactions may result in a rank-related differentiation of social-activity profiles and ‘repayment’ of social acts. The entities are completely identical at the start of the simulation and very simple: they lack a predefined tendency to reciprocate, but are gregarious and perform interactions in which winning is self-reinforcing. They perceive the others’ capacity to win either indirectly, by refering to former experiences with partners, or directly. In this artificial world entities can be ordered according to a dominance hierarchy and a social system with a spatial structure is formed. Furthermore, entities of different rank have their own behavioural profile and patterns of reciprocation emerge. These patterns are particularly apparent in loose groups and among the simplest entities (i.e. those that perceive rank directly). These epiphenomena may also contribute to reciprocation found in real animals such as primates.